摘要 |
PURPOSE: To calculate more accurate bit quantity by employing PIM calculation so as to make the bit quantity constant. CONSTITUTION: This image coding method is made up of following steps, step a) an initial quantization step size is set by a PIM (f) calculation with respect to an image signal string for each image pattern, step b) an image signal string for each image pattern is divided into a plurality of M×N block image signal strings, step c) an assignment bit number to be coded from each block with respect to a prescribed bit rate is obtained by a PIM (bi) calculation with respect to a signal string for each block, a step d) a signal string for each block is coded into a signal string coded by a quantization step size set at present, step e) the bit number of the coded signal string for each block and the assigned bit number are subtracted and the difference is accumulated, and a step f) the current quantization step size is adjusted, when the accumulated value deviates from the preset range. |